The present invention relates to a scanning type exposure apparatus, a position control apparatus, and a method therefor and, more particularly, to a scanning type exposure apparatus which scans to expose a wafer by projecting a pattern of reticle with a strip of illuminating light, a position control apparatus which controls the position of a stage, and a method therefor.
An XY stage is commonly used for determining a precise position in a two-dimensional plane. In a semi-conductor exposure apparatus, an XY stage is used as a wafer stage in order to move a position on the wafer to a reference position. In a scanning type semi-conductor exposure apparatus which scans at a constant speed, it is necessary to control a reticle stage and a wafer stage to be synchronized at high precision when scanning sections to be exposed (referred as "exposure sections", hereinafter). Further, since it is also necessary to keep the luminous exposure constant, it is important to move both stages at a constant speed.
In order to keep the scanning speed constant during scanning of the exposure section, the XY stage is accelerated so that a predetermined conditions for settling (referred as "settling conditions", hereinafter) are satisfied before reaching the exposure section, and the settling conditions have to be kept satisfied during scanning of the exposure section. The settling conditions are determined on the basis of a variation in luminous exposure and other situations. As an example of the settling conditions, in a case where the XY stage moves while its speed is controlled, the deviations between reference speeds and actual speeds of the XY stage has to be within the tolerance for a predetermined period of time.
However, the characteristics of controlling the XY stage depend upon X and Y coordinate values, thus the settling time required for settling by accelerating the XY stage varies from one position on the wafer to an other since transient response also changes in accordance with the X and Y coordinate values. If a reference locus of the XY stage is determined on the basis of the maximum settling time, i.e., since the XY stage is accelerated until it is settled, it takes too much time to scan the entire wafer, which has bad effects in terms of throughput. There is an often used method in which a gain table having coordinate values in the X and Y coordinate system as indexes is used and the gain which is set in a control structure for controlling the XY stage is changed in accordance with the position of the XY stage (indexes). Further, a method adapting an adaptive control is also often used for compensating variations in characteristics of controlling of the XY stage.
However, it is impossible to completely compensate a variation in characteristic for controlling the XY stage only by using the gain table, and variations in settling time, which is the period of time required by the XY stage to satisfy the settling conditions, for different chip of a wafer can not always be compensated. Thus, the aforesaid problem can be partially solved by using the adaptive control and a control system becomes complicated.
FIG. 2 is a block diagram illustrating a configuration of a scanning type exposure apparatus. The function of each block will briefly explained below. Note, the scanning direction is defined as X direction.
A designation value generator 1 determines the X coordinate position 12 and Y coordinate position 10 of the next chip to be exposed on a wafer 8 on the basis of the present position of the wafer 8. Further, the designation value generator 1 also outputs a scanning speed 11. A wafer reference position generator 3 and a reticle reference position generator 2 determine a reference position locus (a locus indicating relationship between time and reference positions) in the scanning direction on the basis of the scanning speed 11 and the X coordinate position 12 of a chip which are inputted from the designation value generator 1. Then, the wafer reference position generator 3 outputs a wafer reference position data 30 to a wafer position controller 5 and the reticle reference position generator 2 outputs a reticle reference position data 20 to a reticle position controller 4. As a result, the wafer 8 and a reticle 7 are controlled to be at reference positions. When the wafer 8 and the reticle 7 arrive at predetermined positions, an exposure light on/off signal 13 is outputted from the designation value generator 1 to a projecting unit 6, and an exposure operation begins. The projecting unit 6 emits a strip of illumination light toward the reticle 7.
FIG. 3A shows an exposing order of chips on the wafer 8. Further, FIG. 3B is a drawing showing a detailed scan cycle in one of the chips shown in FIG. 3A. FIGS. 4A to 4C show responses in position and speed of the chip, and in synchronous error between the chip and a reticle with respect to time when scanning the chip.
A general processing sequence for scanning exposure for one chip is as follow. Note, before starting the processing, the position of the reticle is moved to a predetermined position with respect to the position of the wafer.
(Step 1) The designation value generator 1 reads a settling time from a parameter table which stores characteristics of a position control system. It is assumed that a scanning speed is given.
(Step 2) Next, the chip to be exposed, X and Y coordinates of the start position of scanning the chip on the wafer (simply called "start position", hereinafter), and total amount of displacement to the position where the exposure starts in the scanning direction are determined on the basis of the read settling time and parameters, such as the scanning speed or the current position of the wafer, the maximum acceleration, and so on.
(Step 3) The wafer moves from the current position toward the start position in step movement.
(Step 4) The wafer is accelerated.
(Step 5) Settling operation is performed for a predetermined settling time.
(Step 6) Exposure operation by a strip of illumination light starts after the settling operation.
(Step 7) The exposure operation is performed by displacing the wafer by an amount which depends upon a predetermined chip size and the width of the strip of light.
(Step 8) The exposure operation is completed.
(Step 9) The wafer is decelerated to stop.
(Step 10) The process returns to step 1 when necessary, and the next chip is exposed.
In the aforesaid conventional method, the processing is performed on the assumption that the settling time is constant for scanning each chip (one scanning operation). However, since the position of the XY stage with respect to the apparatus differs in respective scanning operations, characteristics of the position control system changes for each scanning operation. Therefore, a settling time, required in practice, differs in each scanning operation. In other words, although the predetermined settling time is constant, the settling time required in practice changes for each scanning operation. Accordingly, there is a case where an exposure operation may be started before the synchronous errors satisfy the tolerance conditions for exposure. In contrary, there is a case where settling operation is performed for more than a necessary period of time in spite of the fact that the exposure conditions are already satisfied, since the time for initiating exposure (exposure initiation time) has not come yet.
In order to overcome the former problem, it is possible to select and set the longest settling time which is required in practice as a fixed settling time. However, if the settling time is set to a fixed value, the settling operation will continue for an unnecessary period of time and throughput will be decreased.
Further, it is possible to determine a settling time for each scanning operation by actually exposing and evaluating the exposure result in order to determine proper time for initiating exposure. However, such an operation requires many complicated processes, which decreases processing efficiency.